The implications of elections for federalism in Iraq: toward a five-region model.

For most Western observers, Iraq's recent elections represented an unambiguously positive development in the country's traumatic recent history. Despite insurgents' threats to "wash the streets with blood," Iraqis turned out in their millions to participate in Iraq's first democratic elections since the mid-1950s. In some ways, however, the elections created more problems than they resolved. Three issues stand out in this respect. First, despite high turnouts in Kurdish- and Shi'a-dominated areas (extremely high in the case of the Kurds), turnout in Sunni areas was negligible. Second, the triumph of the United Iraqi Alliance (UIA) at the federal level and of various Shi'a religious parties at the provincial level raises serious, potentially divisive questions about the future role of religion in Iraq's political life. Finally, the spectacular electoral performance of the Kurds at both federal and provincial levels has invested the Kurdish leadership with decisive bargaining power over the future trajectory of Iraq's democratic experiment. This is a position they are likely to exploit to maximum effect in the coming months.

Less noticed, but of potentially greater significance, nearly 90 percent of Kurds participated in a simultaneous unofficial referendum on Kurdish independence. Close to 99 percent of those voting voted to secede from the Arab part of Iraq. In sum, the 30 January 2005 elections marginalized almost entirely a significant portion of Iraq's population (Sunni Arabs) while delivering control over the federal government and multiple provincial governments to Shi'a religious parties and simultaneously highlighting the strength of Kurdish secessionist aspirations in the North. As Iraq moves toward stage two of its democratic evolution--the drafting of a permanent constitution--it will require a triumph of constitutional engineering to hold together the state of Iraq within a stable, democratic framework. (1)

The adoption of a federal system for Iraq is a necessary, but not sufficient, condition for achieving this. On this, there appears to be at least a minimum of consensus among Iraqi politicians. However, the precise configuration of such a system and, in particular, how the subunits are to be defined (ethnically or territorially) are issues that have yet to be resolved. Most Western scholars advocate a form of territorial federalism as a defense against the fragmentation of the state, an argument based on solid, plausible theoretical foundations. The political realities of contemporary Iraq mean, however, that this theory prescribes a form of federalism that will ultimately produce precisely the outcome it is designed to avoid. Indeed, we argue that anything other than a form of ethnic (or "plurinational") federalism will result in the disintegration of the Iraqi state. Specifically, we advance an argument for a federal system based on five regions.

The argument proceeds as follows. In the following section we briefly review the literature associated with the design of federal systems, particularly as it relates to the use of federal arrangements to reduce and regulate conflict in deeply divided societies. We then examine how the interethnic and sectarian tensions that have afflicted Iraq since its creation in the 1920s have intensified since the removal of Saddam Hussein's regime and in light of the recent elections. Based on this analysis, we conclude that a federal political system is the only way that Iraq can hope to survive as a coherent territorial entity. We then critically analyze the various proposed blueprints for a federal Iraq and assess the benefits and shortcomings of each. Subsequently, we outline the contours of a federal system based on five regions and illustrate its advantages compared with other proposed federal structures.

WHY FEDERALISM?

One of the major issues on which most Iraqi politicians agree is that Iraq must adopt a federal system in order to survive as a coherent territorial entity. (2) Beyond this, there seems little consensus on how such a system would work in practice or even the meaning of the term "federalism" itself. Most definitions of the term "federation" share common features: at least two levels of government (federal and regional), with separate powers or competencies allocated to each level via a written constitution (though powers may also be shared between levels). In meaningful federations the constitutional division of powers between levels cannot be altered unilaterally by either level of government; the consent of both levels is required (via constitutional amendment). (3) Beyond this, there is little consensus regarding specifics; indeed, almost every aspect of federalism--from definitions, to measurement issues, to effectiveness--is a topic of controversy. (4) Among the most controversial issues is the effectiveness of federations as mechanisms for conflict regulation in divided societies. Several of the world's most durably successful democratic federations--the United States, Germany, Austria, and Australia, for example--have highly homogeneous societies, dominated by a Staatsvolk. (5) In these societies, federal systems may be desirable for any number of reasons, but they are not necessary as a means of holding the state together. In contrast, in societies such as Iraq, with long histories of intercommunal tensions and powerful secessionist sentiments, a federation may be the only way to sustain democracy while maintaining the territorial integrity of the state. Precisely how a federation should be organized to achieve this goal is, however, an open question.

Despite a voluminous extant literature on federalism, issues relating to the character of subunits in federal systems--such as the implications of having fewer rather than more subunits--have been relatively neglected. (6) The geography of subunit boundary delimitation in divided societies has been studied in greater detail, and from this body of work, it is possible to synthesize two basic approaches. The first of these--sometimes referred to as "territorial" (or administrative) federalism--advocates drawing boundary lines to divide rather than unite communities. This "divide-and-conquer" approach results in no single community enjoying majority control in any subunit; each subunit becomes, instead, a demographic microcosm of the state as a whole. (7) By drawing boundary lines to "crack" territorially concentrated communities, the intention is to promote shifting coalitions at the subunit level. In a state with three equally populous, geographically concentrated communities (A, B, and C), for example, the ideal would be for each subunit to contain approximately equal proportions (one-third) of each community and for intercommunity coalition building to become necessary in order to govern. Moreover, the hope is that governing coalitions are fluid over time, such that, today, A governs in alliance with B; tomorrow, B with C; and the day after, C with A; and so on. In sum, this approach avoids majority control at the subunit level such that no community need be permanently excluded from power, and each has powerful incentives to cooperate with others. (8)

A variant on this same basic logic seeks to foster intracommunity competition rather than intercommunity cooperation. The key here is the number of subunits rather than the social composition of their respective populations. The goal is not to create subunits with equal populations of each community but to ensure that the territory occupied by each community is divided into multiple subunits. Horowitz eloquently defends this approach using Nigeria as his example. (9) The transition from a federal system based on three large regions, each dominated by an ethnic majority (the Hausa-Fulani, the Ibo, and the Yoruba), to a nineteen-state model that dispersed rather than concentrated ethnicifies produced "a lively state politics and a more complex--and therefore less tense--politics at the center." (10) Drawing boundary lines that divide up geographically concentrated communities aims at shifting the nature and locus of confrontation from intercommunity conflict at the center to more manageable intraethnic competition at the regional level. In Horowitz's words, "the nineteen states thus created a new, lower layer of conflict-laden issues around which already existing differences crystallized, greatly reducing the previously unchallenged importance of contention at the all-Nigeria level." (11)

At heart, both divide-and-conquer strategies rely on the logic of pluralism. (12) They aim at maximizing the fragmentation of political interests to preclude the emergence of coherent (and potentially tyrannical or inflexible) majority groupings. They differ mainly in that the first seeks to foster cooperation among communities, whereas the latter aims at generating intracommunity competition. Both are proactive strategies that rest on (mostly) unspoken assumptions about the motivations of actors--namely, that the desire to access political power and resources is sufficient to override intercommunal hostilities or intracommunal coherence.

The second major approach, sometimes termed the "ethnic" or "plurinational" model of federalism, rests on a different set of assumptions and generates different policy prescriptions regarding the drawing of boundary lines. (13) This approach seems most appropriate in contexts where the divisions among communities are especially deep and intractable, where secessionist sentiment is intense, or where divide-and-conquer approaches are either logistically unfeasible or rejected by one or more communities. In the aftermath of an ethnic civil war, for example, the creation of defensible "ethnic enclaves" may be the only way bring an end to violence and preclude ethnic cleansing. (14) The immediate goal is the survival of the state as a single territorial entity, so the emphasis is on drawing boundary lines to separate communities, one from the other. The assumption underlying this approach is that there may be occasions on which the intensity of intercommunal hostility is such that cross-communal political appeals are "unlikely to be made and even less likely to be heard." (15) Drawing boundary lines to force communal coexistence may simply exacerbate tensions, leading to further violence. Drawing boundaries to coincide with communal divisions may also be the preferred option under less extreme circumstances. For example, when a minority community fears the cultural, linguistic, ethnic, or religious hegemony of the majority community, intercommunal tension can be avoided by allowing the minority fights of self-government within its own communal enclave. Alternatively, multiple communities within a state may seek to safeguard communal identities while simultaneously enjoying the benefits of participation in a larger, more powerful political unit.

These two "models" of federalism--territorial and plurinational--are ideal types that are nowhere perfectly replicated in the real world. Moreover, the hybrid Canadian system, in which a linguistic minority is largely contained within its own subunit, while anglophone Canada is cracked into nine separate provinces, illustrates that the two are not mutually exclusive. Both have advantages and drawbacks, and which is the preferable option in any given circumstance depends to a large extent on the specifics of the state in question--an issue to which we now turn.

IRAQ IN THE WAKE OF ELECTIONS

The basic contours of Iraq's myriad societal divisions have been extensively documented and are, by now, well known. (16) Historically, the two most significant divisions have been ethnic (Kurd/Arab) and sectarian (Sunni/Shi'a). Currently, Kurds comprise 15-20 percent and Arabs 75-80 percent of the population. The sectarian divide between Sunni (approximately 35 percent) and Shia (approximately 60 percent) (17) sects of Islam is less clear-cut and significantly more complex in its dynamics. (18)

The Impact of Elections

The success of Iraq's 30 January 2005 elections (in terms of voter turnout and the relative absence of violence) was, to a significant extent, overshadowed by the implications of the results themselves. Most notably, turnout among Sunni Arab voters was negligible. In Sunni-dominated Anbar province, for example, recorded turnout for the National Assembly election was 2 percent, and for the provincial council elections, less than 0.5 percent. Estimating Sunni Arab turnout throughout Iraq as a whole is extremely difficult, but based on votes for the only two significant Sunni Arab parties that participated (Ghazi al-Yawer's "The Iraqis" and the Iraqi Islamic Party [IIP]) an educated estimate would be between 5 and 10 percent. (19) Whether through principled opposition to foreign occupation, fear of insurgent reprisals, or unwillingness to participate in a losing venture, the vast majority of Sunni Arabs opted not to exercise their democratic right to vote. The marginalization of Sunni Arabs from the political process (self-imposed or otherwise) and the virtual exclusion of their representatives from the National Assembly create a serious problem for future political stability. (20)

One consequence of the exceptionally low Sunni Arab turnout is that other groups--especially the Kurds--are overrepresented in the Assembly. Table 1 provides data on the vote and seat distribution for parties competing in the January election.

Comprising perhaps 20 percent of Iraq's population, the Kurds turned out in huge numbers to ensure the Kurdistan Alliance--a coalition of, principally, the Kurdistan Democratic Party (KDP) and the Patriotic Union of Kurdistan (PUK)--over 27 percent of the Assembly seats. Kurdish parties also enjoyed disproportionate triumphs at the provincial level, winning majority control in five of Iraq's eighteen provinces. A notable Kurdish victory was scored in the province (Kirkuk) containing the oil rich and deeply divided city of Kirkuk, where a Kurdish-dominated list obtained twenty-six of the council's forty-one seats.

Less noticed, but of greater potential long-term significance, the Kurds also participated in an unofficial referendum on secession from the state of Iraq. According to observers, an equivalent turnout to the national election yielded a "Yes" vote of close to 99 percent. (21) While the Kurdish leadership has consistently stressed its willingness to negotiate and reach consensus within the framework of unified, democratic, federal Iraq, the Kurdish "street" seems increasingly intent on independence for its region. In the short term, this outcome exerts immense pressure on Kurdish leaders not to concede on issues deemed of vital interest to the Kurdish people. (22)

The third electoral outcome that threatens to undermine Iraq's social and political stability is the performance of Shi'a religious parties at both federal and provincial levels. At the national level, the UIA--a coalition of parties that included the Supreme Council for the Islamic Revolution in Iraq (SCIRI), various factions of al-Dawa, and the Badr Brigades (now renamed the Badr Organization)--obtained an overall majority of seats in the National Assembly (140 from 275). The UIA was pieced together with the assistance of Grand Ayatollah Ali Sistani and was openly promoted by Shi'a religious authorities in mosques throughout Iraq. The main secular Shi'a alternative--Iyad Alawi's "Iraqi List"--performed poorly, winning forty seats on an overall vote of less than 14 percent. In the national elections, the UIA won either majorities or pluralities in twelve of Iraq's eighteen provinces, while the Kurds triumphed in five, and Alawi's Iraqi List in just one (Anbar).

The dominance of Shi'a religious parties at the provincial level was, if anything, even more pronounced than at the federal level. SCIRI won the plurality or majority of seats on four provincial councils (Najaf, Karbala, Qadissya, and Muthana), while parties backed by the radical cleric Muqtada al-Sadr won majority control in Wasit province and a plurality of seats in Misan. The hardline al-Fadilah Party (The Islamic Virtue Party (IVP)) (23) won representation on councils in Baghdad (six seats), Basra (twelve), Theqar (eleven), Karbala (five), and Qadissya (three). (24) Overall, Shi'a Islamic parties now control all provincial councils from Baghdad south to Basra.

At the federal level, the policy implications of the UIA's victory are far from clear. Although it is dominated by Shi'a religious parties, the Alliance includes secular Shi'a (such as Ahmed Chalabi's "Iraqi National Congress" (INC)) and representatives of other ethnic groups (Kurds, Turkomans, and even Sunni Arabs), so in practice it is unlikely to function as a coherent voting bloc. (25) The likely outcome is political fragmentation at the federal level, leaving the responsibility for the day-to-day governance of Iraq in the hands of the newly legitimized provincial councils. The danger, therefore, is that Shi'a religious parties will have their greatest impact at the local, rather than federal, level and that by the time a permanent constitution is approved and the new government takes office (by December 2005), the southern provinces will have evolved into a constellation of Islamic statelets policed by religious militias. (26) Of equal concern, the electoral triumph of religious parties will greatly complicate efforts to define the role of Islam in the permanent constitution. Efforts to craft compromise language during the writing of the Transitional Administrative Law (TAL) were nearly derailed at the eleventh hour by a conflict over whether Islam should be the source of law or merely a source of law; this dispute was never definitively resolved. (27)

As Iraqis begin the process of crafting their permanent constitution, these three, related developments--the political marginalization of Sunni Arabs, the empowerment of the Kurds, and the emergence of Shi'a religious parties as the country's dominant political force--will need to inform the process. Institutions need to be designed to help reintegrate Sunnis into the political process (thereby isolating the insurgency from its support base), (28) to constrain the influence of Islamic parties, and to placate the Kurds (thereby defusing secessionist sentiment).

FEDERAL OPTIONS FOR IRAQ

Of the various options outlined above, one can be eliminated immediately. Among scholars pondering the design of Iraq's future political system, there is little enthusiasm for drawing boundary lines to guarantee that no communal group enjoys a majority in any province. (29) There are two, straightforward reasons for this. First, given Iraq's population breakdown, and the geographical distribution of its communities (see Figure 1), it is mathematically and logistically impossible to create such provinces.

With 60 percent of Iraq's population, Shi'a Arabs would inevitably constitute majorities in many provinces (up to one half). The only region of Iraq where this approach is feasible is in the north, where the elongation of the three Kurdish-dominated provinces could capture enough Sunni Arabs to deny the Kurds an outright majority in any province. Iraq's ethnic Turkoman population, geographically sandwiched between Sunnis and Kurds, might then hold the balance of power in these regions. This leads to the second problem with this approach. From a Kurdish perspective, it is unacceptable. (30) Rather than submit to the redrawing of boundaries in this way, the Kurds would almost certainly withdraw from Iraq. (31) For this to be a permanent option, then, it would have to be imposed on the Kurds by force, and the only military power capable of doing this at present is the United States.

The second divide-and-conquer approach--carving up Iraq's communities into smaller subunits to encourage intracommunal competition--offers more promise and is overwhelmingly the approach of choice for the majority of scholars pondering the future of Iraq. With few exceptions, (32) the scholarly consensus favors either a vaguely defined "territorial" (in preference to ethnic) federal system, or a system based on Iraq's already existing eighteen provinces. (33) As indicated by Figure 2, the existing system divides the Kurdish region into three (or four) provinces, Sunni Arab Iraq into three (or four), and Shi'a Arab Iraq into nine. Hence, advocates of retaining Iraq's existing eighteen provinces are, implicitly or explicitly, relying on the logic of the second divide-and-conquer approach. (34)

One immediate benefit of this approach is that it can be implemented within existing boundaries, thereby avoiding endless, potentially divisive debates over where to draw the lines. Second, and more important, if the goal is to fragment communities, thus fostering intracommunal competition and transferring conflict from the center to the regions, this option stands a good possibility of success, especially in the Kurdish-dominated north. Currently, the two major Kurdish factions--the KDP and the PUK--each control one province in its entirety. The KDP dominates in Dohuk and the PUK in Sulaymaniyah, leaving Erbil as a political (and, at times, military) battlefield. During the Kurds' "golden era" (1992-2003), the Kurdish region was politically and physically divided between the PUK and KDP for most of the period. In 1996, the two parties fought a brief but bloody war for control of Erbil city and the respective leaderships of the two have a long history of mutual hostility. It would not require much to spark renewed hostilities between the two. If the goal is to diminish the political power of the Kurds as an ethnic group and to deflate their secessionist ambitions via the promotion of intra-ethnic conflict, then maintaining existing boundaries may well prove an effective strategy. In the south, too, it is arguable that the eighteen-province model is an effective way to check the power of Shi'a religious parties. Unlike at the federal level, where religious parties competed as part of the UIA coalition, at the provincial level, these parties for the most part ran independently. Thus, in most provinces, religious parties competed for power rather than collaborating. In only two southern provinces--Wasit and Karbala--did single Islamic parties or groupings win majority control over provincial councils. In most others, the vote for Islamic parties was highly fragmented, yielding divided councils. Assuming that a fragmented, internally competitive Shi'a Islamic movement is preferable to a single, unified Islamic bloc with a consolidated power base in the south, then the evidence for maintaining the nine provincial divisions as they currently exist is persuasive.

In terms of reintegrating Sunni Arabs into the political process, it is not clear that the logic of the eighteen-province model has much relevance. Iraq's Sunni population is already highly fragmented and incapable of acting as a coherent bloc. This applies as much to the Sunni insurgency as it does to political actors. The only credible Sunni Arab party to participate in the elections, the IIP, gained only 21,000 votes at the federal level and obtained no seats in the Assembly. The IIP did "win" control of the Anbar provincial council, if one can speak of "winning" an election in which less than 0.5 percent of the population turned out to vote. The only Sunni Arab "mouthpiece" to have emerged since the removal of Saddam Hussein's regime is the Association of Muslim Scholars (AMS), an organization that is implacably hostile to the presence of foreign forces and that considers the newly elected government to be the illegitimate product of an illegitimate process. It is doubtful that a federal system based on eighteen provinces can do anything to affect this situation.

Overall, though, the logic of constructing a federation on the basis of the eighteen-province model is plausible (at least superficially), which is probably why it continues to be the option of choice for most analysts. Unfortunately, however, regardless of how convincing the logic, this model is not feasible in practice. There are several possible flaws with this design-not the least of which is that promoting intra-ethnic conflict in the one region of Iraq (Kurdistan) that has remained stable and secure since 2003 is a highly dangerous strategy and one that risks increasing the involvement of external powers (Iran and Turkey) in the internal affairs of Iraq. More importantly, like the previous option, the eighteen-province model is unacceptable to the Kurds. (35) The two Kurdish leaders (Jalal Talabani of the PUK and Masoud Barzani of the KDP) have presented a united front during negotiations over Iraq's political future and have made it clear that the nonnegotiable price for their reintegration into the state of Iraq is constitutional recognition of a distinct Kurdish region. The implication of the Kurdish position is that if Iraq is to survive as a unified territorial entity, then some form of plurinational federalism is inevitable.

Plurinational Federalism: The Options

The inevitability of an ethnically defined Kurdish region as a constituent element in any future federal system does not end the debate. There is still the issue of how to organize the remainder of Iraq. Among the options are a bicommunal federation and a mixed system, such as that outlined in the TAL.

The only selling point of the bicommunal option is that it is acceptable to the Kurds--otherwise, it is deeply problematic. (36) The initial (September 2002) Kurdish proposal for federalism in a future Iraq--contained in a draft "Constitution of the Federal Republic of Iraq"--was a system based on two regions, a Kurdish region with extended borders and an Arab remainder. The Kurds swiftly abandoned this proposal once it became evident that its provisions were unacceptable to Iraq's Arab population and that a bicommunal federation was not the only way to achieve their major objective (a unified Kurdish region within a federal structure). (37)

The TAL, signed in March 2004, functions as Iraq's interim constitution until the permanent constitution is approved in a national referendum. It outlines a compromise position, apparently acceptable to both Kurds and Arabs (those at least that were involved in the drafting process). Article 9 recognizes the status of both Arabic and Kurdish as official languages of state, while Article 53 specifically recognizes the Kurdistan Regional Government (KRG) as the official government of the territory under Kurdish control as of March 2003. Both articles suggest a bicommunal form of federation for Iraq. Article 4, however, specifically forbids a federal system based upon "origin, race, ethnicity, nationality, or confession," in favor of a system based on "geographic and historical realities," a formula that could be interpreted in various ways. (38) The fact that separate elections took place in all eighteen provinces and the Kurdish region in January 2005 suggests that the TAL appears to envisage a separation of powers between the federal government and the eighteen provinces and, simultaneously, a distinctive relationship between the Kurdistan region and the center. To confuse matters further, the Kurdish region is identified as the three Kurdish dominated provinces (Erbil, Sulaimanya, and Dohuk) but also parts of the territories of three other provinces (Ninevah, Diyala, and Kirkuk). These provinces will, presumably, have a relationship with the center based on their status as provinces and a distinctive relationship based on their incorporation (in whole or in part) into the Kurdish region.

The TAL, therefore, creates two tiers of federalism--a relationship between the center and the eighteen provinces and another between the center and the Kurdish region--but it is unclear how such a system would work in practice. Assuming that laws passed by the KRG will have supremacy over those passed at the provincial level, this will leave the Kurdish population of these three provinces governed by a different set of laws from the remainder of the population. This problem of overlapping jurisdictions matters only to the extent that the KRG is assigned powers in its relations with the central government that are denied to the eighteen provinces. But this is almost inevitable. The Kurds gain nothing from having a regional government that has power equal to, or less than, individual provinces. Hence, another problem with this proposed arrangement is that Iraqi citizens who reside in the Kurdish region will have privileges, rights, and protections vis-a-vis the central government that citizens in the remainder of Iraq will not have. In effect, the TAL establishes a federal system that is constitutionally asymmetrical in that the Kurds are granted a form of autonomy and constitutional status distinct from the rest of Iraq. (39) Granting constitutional privileges to one group of Iraqi citizens primarily on the basis of ethnic identity will be a difficult position for the Kurds to sell to other groups in Iraq. The major political party in the National Assembly, the UIA, campaigned on a platform that called for an "Iraq that provides a climate of peaceful coexistence among Iraqis without preferential treatment for any group." (40) This campaign pledge is difficult to reconcile with the existence of a Kurdish region that will, inevitably, involve "preferential treatment" for one of Iraq's groups. This highlights a more fundamental tension between the Shi'a and Kurds that cannot help but create friction during negotiations over the permanent constitution. The concern on the part of Shi'a groups (and Grand Ayatollah Sistani) to ensure constitutional equality for all Iraqis, regardless of ethnicity or sect, conflicts with the Kurdish position that they deserve special treatment because of their ethnicity. This tension is clearly evident throughout the TAL. The common assumption that the TAL grants Kurds veto power over the permanent constitution is, in practice, true but technically false. The relevant article (Article 61) requires majority approval across Iraq and that "two-thirds of the voters in three or more governorates do not reject it." (41) Thus the TAL awards veto power to any three governorates (provinces), not just the three dominated by the Kurds. This formula explicitly avoids granting the Kurds a power that is denied to other groups in Iraq. This tension also helps explain why the federal system outlined in the TAL is ambiguous to the point of incoherence. Resolving this basic tension is key to producing a permanent constitution that is acceptable to most, if not all, parties. A federal system that allows the Kurds to form a powerful region from three (or four) provinces but that denies this right to other Iraqi groups is a recipe for confrontation. (42) A nonnegotiable demand on the part of the Kurds is unacceptable to Iraq's Shi'a (and Sunni) population. One plausible solution to this dilemma is inspired by Article 52 of the TAL, according to which any three provinces "shall have the right to form regions from amongst themselves." A federal system based on regions, while not ideal, resolves many of the problems associated with the alternative models. Most important, it is potentially acceptable to a broader range of political forces.

The five-region model championed by, among others, Iraq's national security advisor Mowaffak al-Rubaie envisages the creation of the following regions: Basra province (to include Basra, Nasariyyah, and Amara), Kufa province (to include Karbala, Najaf, Kufa, and Hilla), Greater Baghdad (to include Ba'quba), and Mosul province (to include Mosul, Tikrit, Fallujah, and Ramadi). Combined with the established Kurdish region in the north, these five regions would form the basis of Iraq's new federal system. (43) Figure 3 illustrates one possibility for where the boundary might fall in such a system.

To avoid preferential treatment for any group, the five regions should enjoy constitutional symmetry. The powers granted to one should be granted to all. The Spanish model, whereby all regions have the option of negotiating directly with the federal government to extend autonomy beyond that granted in the constitution, should also be seriously considered for Iraq. (44)

Benefits of a Five-Region Model

In addition to constitutional equality, these five regions would have rough equality in terms of population. Under any system that preserves the eighteen provinces (or fifteen plus a Kurdish region), Baghdad, with approximately 7 million inhabitants, dwarfs all others in term of population. One of the purposes of adopting a federal system for Iraq is to devolve power from the center to the periphery (a point of consensus among scholars), but this is difficult to achieve when the center is simultaneously the capital city, a province, the cultural, economic, and communications hub of the country, and home to over one-fourth of the country's population. The five-region model creates regions with a more symmetrical population distribution. Greater Baghdad would still be the most populous, but it would coexist with four other regions, each with 4-5 million inhabitants. That five regions will have more power (all else being equal) relative to eighteen provinces is largely self-evident. There are fewer of them and each represents a larger proportion of territory, resources, and population, thus enhancing their political influence in relation to the central government. The importance of this extends beyond the devolution of power. Historically, the extreme level of centralization that defined the Iraqi political system has made the country vulnerable to military coups. On several occasions, relatively minor forces have been able to effect regime change by establishing control over the capital city. The establishment of five power centers in Iraq will make this more difficult.

By far the major benefit of this model, however, is that it is potentially more broadly acceptable to major political forces than any alternative. From the Kurdish perspective the major advantage of the regional model over all other alternatives is that it is a compromise position that allows the Kurds to keep the autonomy they already enjoy but that is also likely to prove acceptable to other Iraqi groups, particularly the Shi'a. According to an aide of Ayatollah Sistani, "the creation of two main provinces for Shiites, one with oil and another one with the clerical schools and shrines is exactly what Sistani would like to see now." (45) If the Kurds push for a regional model, therefore, they will find themselves allied with Iraq's most influential political figure, rather than opposed to him as they are now. Likewise, Ahmed Chalabi, the secular leader of the Iraqi National Congress (a faction within the UIA), has argued for the establishment of a southern region in any future federal system, claiming, "The people of the south want a southern region as part of the federal union of Iraq. (46) Elsewhere, a top official of SCIRI has called for the establishment of a Holy Region centered on Karbala, and the Sunni Arab governor of Anbar province has proposed a large Sunni region, comprising Saladin, Ninevah, and Anbar provinces. (47) The five-region model is, therefore, a compromise position that allows the Kurds to keep their autonomy but that does not involve "preferential treatment for any group" and is potentially acceptable to many of Iraq's major political forces.

It is unclear how Iraq's broader Sunni Arab population would react to a five-region proposal. However, absent the restoration of Sunni control over a powerful central government, the regional model is likely to be at least as acceptable as any alternative. It will allow Sunnis to govern and police themselves in the Mosul province, as well as to maintain a position of (possibly decisive) influence in Greater Baghdad. If nothing else, it will protect Sunnis against Shi'a dominance and the possibility that the Shi'a will exploit their newly acquired position of power to exact retribution for decades of oppression. Reintegrating the Sunni into the political process and ending the insurgency is impossible at present because there is no coherent Sunni leadership with which to negotiate. The forces currently uniting Sunni Arabs are hostility toward the occupation and resentment at the new political order beyond this, they are hopelessly fragmented as a political force. The widespread nonparticipation of Sunni Arab voters in the election guaranteed the virtual exclusion of their representatives from the National Assembly. It also gifted control of key provincial councils to the Kurds. In particular, Kurdish control of Ninevah, and, thereby, the city of Mosul, is likely to exacerbate already deep Kurd/Arab divisions in the city. There are no easy solutions to the problem of Sunni marginalization. However, a single, unified, Sunni-dominated region (to include Mosul) allows for the possible emergence of a broad, catchall political force that represents, and can negotiate on behalf of, Sunni Arab interests. This is preferable to the prevailing situation in which Sunni interests are fragmented among numerous provinces, several of which are now controlled by Kurds.

Restraining the growth in power of Shi'a religious parties is the third major test of the new federal system in Iraq. As noted above, the immediate danger is at the provincial rather than federal level. The UIA is dominated by religious parties--principally, al-Dawa and SCIRI, both of which have the stated goal of creating an Islamic state in Iraq, but this loose coalition of interests enjoys a majority in the National Assembly only because of the participation of secular Shi'a factions (such as Chalabi's INC). Moreover, the new government is likely to do very little governing between now and the election of the first government under the permanent constitution. For the foreseeable future, the day-to-day governance of Iraq will fall to the provincial level by default, and it is here that religious parties made their most impressive gains. Keeping the south of Iraq divided into nine provinces, many with relatively small populations, appears the logical way to ensure that the Shi'a religious community remains internally fractured. This is one approach, but not necessarily the best. As with the Kurds, there is a real danger that violent conflict rather than healthy political competition will result from any divide-and-conquer approach. All of the major religious factions have attendant militias that have been known to engage in violent conflict to gain political control over certain cities. (49)

Based on the results of the recent elections in the South (provincial and national), there is evidently broad popular support for religious parties. (50) Isolating these forces is, therefore, not an option--nor will it be until a viable secular alternative emerges. Until this happens, the emphasis must be on providing structural incentives to promote moderation.

The small size of many southern provinces in terms of population means that it requires a relatively small number of votes to establish control over provincial councils. For example, the southern province of Muthana has a population of just over 500,000. Based on turnout for the recent provincial election (approximately 123,000), a seat on the provincial council can be gained with only 3,000 votes. Theoretically then, 66,000 votes are sufficient to obtain majority control of the council. This favors highly motivated parties--such as the IVP--that may enjoy only limited popular support but that are able to mobilize effectively the support they have. All else being equal, the electoral impact of these parties is diluted the larger the electorate, which suggests fewer, larger regions in the south may be preferable to a larger number of lightly populated provinces. Larger regions will also require parties to form broad electoral coalitions or risk being excluded from power. Insofar as coalitions promote values of compromise and consensus, they tend to have a moderating influence on behavior. Hard-line religious parties have the option of isolation or moderating their positions to participate in catchall coalitions.

Most importantly, the creation of a Holy Region that encompasses Iraq's two holiest cities--Najaf and Karbala (and Kufa)--offers something of a compromise between those seeking a wholly secular Iraq and those advocating a greater role for Islam in Iraq's political life. The creation of a Holy Region governed according to the "universally accepted tenets of Islam" will likely diminish demands for these to be incorporated into Iraq's constitution, thereby removing a major point of tension among Iraq's communities. The status of Islam as either a or the source of law need not be determined at the federal level. Rather, each region can be permitted to apply the tenets of Islam (or not) at its own discretion. This will not produce a Holy Region governed according to Western, liberal democratic standards, but this is unlikely to happen in any case.

Drawbacks of the Five-Region Model

The two most obvious objections to the five-region model are, first, that the division of Iraq into five separate communal "enclaves" will entrench rather than ameliorate intercommunal tensions and, second, that this model is a recipe for the future fragmentation of the state. The first of these objections speaks to the broader theoretical debate regarding the most effective institutional means of dealing with divided societies. Political institutions can be crafted either to reflect or to deny the significance of Iraq's communal divisions. While the significance of communal divisions cannot be quantified in any meaningful way, certain historical trends are undeniable. Since Iraq's creation in the 1920s until the overthrow of Saddam Hussein, Sunni Arabs governed and Shi'a Arabs and Kurds were systematically excluded from political power. Since 2003, Iraq has been governed by the Iraqi Governing Council, (51) with seats allocated on a communal basis, and an interim government, (52) with a secular Shi'a prime minister, a Sunni Arab president, and a religious Shi'a and a Kurd as the two vice presidents. Iraq's first democratically elected government perpetuates this tradition, with ministerial portfolios allocated on a communal basis, and a Presidential Council comprising a Kurd (President Jalal Talabani), a Sunni Arab (Vice President Ghazi al-Yawer), and a religious Shi'a Arab (Vice President Abdul al-Mahdi). Both the election and the ongoing insurgency provide strong evidence that communal identity is now a key factor in Iraqi politics. Sunni Arabs largely boycotted the election, Shi'a Arabs voted in massive numbers for the main Shi'a ticket, and Kurds voted almost exclusively for the Kurdish list. (53) The insurgency is dominated by Sunni Arabs, while Kurds and Shi'a staff the nascent Iraqi defense forces. Communal divisions have always existed (and been exploited) in Iraqi politics, and they have only hardened since the demise of Saddam Hussein's regime. Power sharing on the basis of community identity is already an established fact of political life in Iraq's new democracy because governing on any other basis seems not to be possible at present.

The second objection, that the five-region model would precipitate the fragmentation of Iraq, either through civil war or secession, is valid to the extent that a viable alternative exists that offers better prospects of holding Iraq together democratically. The eighteen-province model--the main alternative--is unacceptable to the Kurds and its adoption would make Kurdish withdrawal from Iraq a virtual certainty. Moreover, the eighteen-province model, while not the cause of intercommunal violence, appears to have done little to assuage it. An Iraq governed under this model has witnessed a steady increase in intercommunal violence since 2003, notably in those provinces and cities with the most diverse populations. (54) Part of the theoretical rationale for maintaining the eighteen provinces is to retain diverse subunits in which communities are "forced" to cooperate in order to govern. The two most appropriate testing grounds for this theory are the provinces of Kirkuk and Ninevah, both of which are characterized by a high degree of ethnic diversity, and both of which have suffered some of the worst interethnic violence in Iraq. In Mosul and Kirkuk cities, initial efforts to generate interethnic cooperation through the use of power-sharing mechanisms swiftly disintegrated, and the recent election has gifted control over both to the Kurds. Kurdish political dominance over the staunchly Sunni Arab city of Mosul and the deeply divided, oil-rich city of Kirkuk can only exacerbate interethnic tensions. (55) While it is unreasonable to attribute increasing levels of interethnic and intersectarian violence to the design of political institutions, the logic on which the eighteen-province model relies is seriously undermined by the absence of intercommunal cooperation in those regions where it should be most evident, and where it is most required. The five-region model separates communities territorially, but it separates communities that are already on the cusp of civil war.

CONCLUSION: COHERENCE OR FRAGMENTATION?

If Iraq is to survive as a single territorial unit, a federal system that includes a Kurdish region is inevitable. Anything less will lead to a Kurdish withdrawal from participation in the institutions of the Iraqi state and, almost certainly, to the Kurds' ultimate secession from the state itself. This basic reality limits the options available for a future federal system. Assuming the inevitability of a Kurdish region, the core issue becomes how to organize the non-Kurdish part of Iraq, and here the basic choice is between fewer, larger regions and a larger number of smaller provinces. All else being equal, the former will encourage the emergence of broader, more coherent political forces, while the latter will promote political fragmentation. (56) In theory, the fragmentation of political forces should generate intercommunal cooperation or, more plausibly, intracommunal competition. In turn, this should make it more difficult to assemble and sustain governing coalitions (as is plainly the case in the National Assembly), thus leading to unstable governments but stable policy (in the sense that it is difficult to create the consensus necessary to change the status quo). In certain contexts, such as the well-established democracies of Western Europe, such an outcome may be desirable. In the context of contemporary Iraq, where a violent insurgency continues to wreak havoc, and where physical, economic, and social reconstruction has barely begun, there is a desperate need for political cohesion and stable governance rather than fragmentation and instability. The five-region model is not the perfect federal system for Iraq. It is, however, the least bad option when ranged against the available alternatives, which is the appropriate standard against which to judge it.

The debate over the design of Iraq's federal system also yields valuable insights regarding the limitations of theory when juxtaposed against political realties on the ground. The ideal solution theoretically--a system that promotes intercommunal cooperation or intracommunal competition--turns out to be unfeasible in practice. The reality of contemporary Iraq is that the Kurds have both the determination and the capacity to secede (or at least withdraw) from Iraq if their demands are not satisfied. Hence, regardless of theoretical plausibility, any argument that posits a federal system that does not recognize the existence of a Kurdish region as its starting point runs the risk of precipitating precisely the outcome it is designed to avoid.

AUTHORS' NOTE: We wish to express our gratitude to three anonymous reviewers, Brendan O'Leary, and Jackie Ravine for their insightful and constructive suggestions on how to improve the quality of this article.

Liam Anderson

Wright State University

Gareth Stansfield

Exeter University

(1) We do not mean to imply from this any normative judgment about the merits or otherwise of Kurdish secessionism or the political dominance of Islamic parties. Indeed, given the treatment the Kurds have received at the hands of successive regimes in Iraq (and Turkey), there is a strong moral argument to be made in favor of an independent Kurdish state in northern Iraq. Likewise, insofar as the success of religious parties in recent elections reflects the will of the majority of the Iraqi people, it is a democratic outcome that is as normatively acceptable as any other. Our point, however, concerns the implications of these developments for the stability of Iraq and its prospects of holding together as a coherent territorial entity. This does assume, of course, a preference for a unified Iraq, and that all groups (including the Kurds) are better off having their preferences satisfied within the borders of a single state rather than suffering through an uncertain and potentially violent process of fragmentation.

(2) The Kurds remain the most vocal advocates of a federal system (based around an ethnically defined Kurdish region). Among Shi'a groups, the UIA's campaign platform for the January elections included a commitment to a "constitutional, pluralistic, democratic and federally united Iraq." This commitment reflects other statements by Shi'a leaders such as the 9001 "Declaration of the Shi'a," a document authored by prominent Shi'a exiles, in which the authors called for a "federal system" in order to "fully accommodate Iraq's diversity" (http://www.al-bab.com/arab/ docs/irag/shia02a.htm). Among Shi'a forces, only radical cleric Muqtada al-Sadr has expressed strong reservations about federalism in Iraq (Nimrod Raphaeli, "Moqtada Al-Sadr: The Young Rebel of the Iraqi Shi'a Muslims," The Middle East Media Research Institute, 11 February 2004, pp. 3-4). The attitude of Sunni Arabs toward federalism is difficult to discern because Sunni Arab political interests are extremely fragmented and there is currently no credible "mouthpiece" of the Sunnis to articulate a position on the issue.

(3) Brendan O'Leary, "Multi-national Federalism, Federacy, Power-Sharing, and the Kurds of Iraq" (keynote address, "Multi-Nationalism, Power-Sharing and the Kurds in a New Iraq," Cafritz Foundation Conference Center, George Washington University, 12 September 2003).

(4) Important works that examine such problems include Daniel J. Elazar, Exploring Federalism (Tuscaloosa: University of Alabama Press, 1987); Ronald L. Watts, Comparing Federal Systems (Montreal and Kingston: School of Policy Studies, Queen's University, 1999); Preston King, Federalism and Federation (Baltimore, MD: Johns Hopkins University Press, 1982); Mikhail Filipov, Peter C. Ordeshook, and Olga Shetsova, Designing Federalism: A Theory of Self-Sustainable Federal Institutions (Cambridge: Cambridge University Press, 2004); Thomas Franck, Why Federations Fail: An Inquiry into the Requisites for a Successful Federation (New York: New York University Press, 1966); Valerie Earle, Federalism: Infinite Variety in Theory and Practice (Itasca, IL: F.E. Peacock, 1968); Ivo Duchacek, Comparative Federalism: The Territorial Dimensions of Politics (New York: Holt, Rinehart and Winston, 1970); William H. Riker, Federalism: Origin, Operation, Significance (Boston: Little, Brown, 1964). For a useful summary article of the state of the discipline, see Daniel J. Elazar, "International and Comparative Federalism," PS: Political Science and Politics 26 (June 1993): 190-195.

(5) A Staatsvolk can be defined as a single national (or ethnic) people that dominates a state demographically and electorally. On the significance of a Staatsvolk for the stability and success of federations, see Brendan O'Leary, "An Iron Law of Nationalism and Federation? A (Neo-Diceyian) Theory of the Necessity of a Federal Staatsvolk, and of Consociational Rescue," Nations and Nationalism 7 (2001): 273-296.

(6) On the importance of the character of subunits, see Ronald Watts, Multicultural Societies and Federalism (Ottawa: Royal Commission on Bilingualism and Biculturalism, 1970).

(7) Arend Lijphart, Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries (New Haven, CT: Yale University Press, 1999), pp. 195-197.

(8) For a summary of the argument endorsing the drawing of boundary lines across communities, see Maurice Vile, "Federation and Confederation: The Experience of the United States and the British Commonwealth," Political Cooperation in Divided Societies, ed. P. Rea (Dublin: Gill and Macmillan, 1982), pp. 216-228.

(9) Donald L. Horowitz, Ethnic Groups in Conflict (Berkeley: University of California Press, 1985).

(10) Ibid., 604.

(11) Ibid., 605.

(12) This is perhaps not surprising given that some scholars consider federalism to be a subset of pluralism. See King, Federalism and Federation, p. 19.

(13) We use the neutral term "plurinational" in preference to the more pejorative term "ethnic" because the latter is usually employed by critics of using federal arrangements to accommodate diverse minorities. We are grateful to an anonymous reviewer for pointing this out.

(14) Chaim Kaufmann, "Possible and Impossible Solutions to Ethnic Civil Wars," International Security 20 (Spring 1996): 136-175.

(15) Ibid., 137.

(16) Standard works on the modern history of Iraq include Charles Tripp, A History of Iraq (Cambridge: Cambridge University Press, 2000); Phebe

Marr, The Modern History of Iraq (Boulder, CO: Westview Press, 1985); Liam D. Anderson and Gareth Stansfield, The Future of Iraq: Dictatorship, Democracy, or Division (New York: Palgrave Macmillan, 2004); Sandra Mackey, The Reckoning: Iraq and the Legacy of Saddam Hussein (New York: W.W. Norton & Co., 2002). On the use of violence as a tool of state policy, see Kanan Makiya, Republic of Fear: The Politics of Modern Iraq (Berkeley: University of California Press, 1989); Said K. Arburish, Saddam Hussein: The Politics of Revenge (New York: Bloomsbury, 2000). For concise summaries of historical divisions, see Dawn Brancati, "Can Federalism Stabilize Iraq?" Washington Quarterly 27 (Spring 2004): 8-11; Andreas Wimmer, "Democracy and Ethno-religious Conflict in Iraq," Survival 45 (Winter 2003-2004): 113-119.

(17) Estimates of the numerical presence of various communities in the population can only be considered rough approximations and the numbers are often hotly contested by the communities themselves. The most recent reliable census was conducted during the last days of the monarchy in 1957. Censuses conducted under Ba'athist rule were highly politicized affairs, often designed to maximize the Arab population at the expense of other ethnicities (particularly the Kurds) in order to serve strategic purposes, such as ensuring an Arab majority in the disputed city and province of Kirkuk.

(18) Geographically, the Kurds dominate in the three northern provinces of Dohuk, Irbil, and Suleimaniyah. They are also a majority (or at least a plurality), in Kirkuk province, home to the oil-rich city of Kirkuk. Elsewhere, Kurds form sizable minorities in Diyala and Ninevah. Sunni Arabs constitute the majority in at least two provinces (Anbar and Saluhadin) and, probably, pluralities in Ninevah and Diyala. All provinces south of Baghdad are dominated by Shi'a Arabs, though many contain significant Sunni minorities. Baghdad itself is a diverse city in which Shi'a Arabs constitute at least a plurality, with significant populations of Sunni Arabs and even Kurds.

(19) Of a total of 8.5 million votes cast in the elections, "The Iraqis" obtained 151,000 votes and the Iraqi Islamic Party, 21,000 votes. This would put Sunni turnout at about 2 percent of the total. Assuming Sunni Arabs comprise 15-20 percent of Iraq's population, this implies that turnout among Sunnis was somewhere between 5 and 10 percent.

(20) At present, Sunni Arabs make up only seventeen of the 275 deputies in the National Assembly.

(21) According to the Kurdistan Referendum Movement, of the 1,973,412 people who participated in the referendum 1,953,762 voted for independence from Iraq (http://www. indybay.org/news/2005/1720506.php; accessed 9 February 2005). These figures should be treated with some caution as they have not been independently verified.

(22) It is possible, but unlikely, that the strong showing of Kurdish parties may have the opposite effect in that the political strength of the Kurds relative to other groups may allow them to secure their interests within Iraq, thus diffusing secessionist sentiment. The problem, however, is that many Kurdish demands are simply incompatible with what Arab groups (who still, collectively, constitute about 80 percent of the population) are likely to be prepared to accept. Unless the Kurds are willing to back down on, say, the status of Kirkuk, it is difficult to see how a major confrontation can be avoided. The threat to the stability of Iraq, then, is that the relative political strength and coherence of the Kurds, combined with the referendum result, will make the Kurds less, not more, willing to compromise on divisive issues such as Kirkuk.

(23) The IVP is sometimes assumed to be the political vehicle of Muqtada al-Sadr. However, while it is part of a broader "Sadrist" movement, the party's allegiance is to Muqtada's father (Muhammed Sadiq al-Sadr) via Ayatollah Muhammed al-Yacoubi, the IVP's founder. In terms of hostility toward the occupation and commitment to an Islamic state in Iraq, however, the IVP and Muqtada are largely indistinguishable.

(24) Of the eighteen councils, seventeen have forty-one total seats, and the other (Baghdad) has fifty-one.

(25) Indeed, Grand Ayatollah al-Sistani's influence may be the only "glue" that currently holds the UIA together. Otherwise, the UIA is a loose coalition of twenty-three diverse (mainly Shi'a) groupings that embrace a wide variety of ideological viewpoints (Ashraf Khalil, "Continued Unity of Shiite Alliance Faces Challenges," The Seattle Times, 30 January 2005).

(26) For example, the armed wing of SCIRI (the Badr Brigades) is the major security force in several southern provinces, such as Karbala and Muthana. Religious militias have also begun to assert themselves in Basra. See George Packer, "Letter from Basra: Testing Ground," The New Yorker, 15 March 9005; http://www.newyorker.com/printable/?fact/050228fa_fact; Solomon Moore, "Recent Violence Stirs Sectarian Tensions in Once-Quiet Basra," The Los Angeles Times, 20 April 2005; http://www.latimes.com/news/nationworld/world/la-fg.basra20apr20,3731259. story?coll=la.home.world.

(27) Once again, the wording of the TAL does not provide much clarity. The agreed wording, that Islam is to be a source of law, sits uncomfortably with Article 7, which forbids laws that contradict the "universally agreed tenets of Islam" (whatever these may be).

(28) A detailed analysis of the political significance of the insurgency is beyond the scope of this paper. Needless to say, unless Sunni Arabs can be reintegrated into the political process, the prospects for defeating the insurgency are bleak.

(29) O'Leary, "Multi-national Federalism," p. 10.

(30) Author interviews with PUK and KDP political leaders, March-April 2004. See also O'Leary, "Multi-national Federalism," p. 11; International Crisis Group, Iraq's Kurds: Toward an Historic Compromise? (Amman/Brussels: International Crisis Group, 2004).

(31) The Kurds would probably stop short of an outright declaration of independence and would instead withdraw behind the "green line" that separates Kurdish from Arab Iraq and govern themselves as a de facto independent state.

(32) A notable exception here is O'Leary, "Multi-national Federalism." See also Karen Eklund, Brendan O'Leary, and Paul R. Williams, "Negotiating a federation in Iraq," The Future of Kurdistan in Iraq, ed. Brendan O'Leary, John McGarry, and Khaled Salih (Philadelphia: University of Pennsylvania Press, 2005), pp. 116-142, for an important discussion of the various possibilities (including a five-region model) for organizing a federal system in Iraq.

(33) Iraq's eighteen provincial boundaries were established in 1968 by the Ba'athist regime, and (with some gerrymandering) have remained in place ever since.

(34) Some scholars advocate territorial federalism on the grounds that anything is preferable to an ethnic form of federalism for Iraq. See, e.g., Kanan Makiya, "A Model for Post-Saddam Iraq," Journal of Democracy 14 (July 2003): 5-12. Others are more explicit about the need to retain the eighteen-province model as a divide-and-conquer mechanism. See, e.g., Adeed Dawisha and Karen Dawisha, "How to Build a Democratic Iraq," Foreign Affairs 82 (May-June 2003): 36-50; Daniel Byman, "Constructing a Democratic Iraq," International Security 28 (Summer 2003): 47-78; Andreas Wimmer, "Democracy and Ethno-religious Conflict in Iraq"; Alec Walen, "Federalism for Postwar Iraq," Findlaw's Writ (http://writ.news.findlaw.com/commentary/20030410_walen.html).

(35) Author interviews with PUK and KDP political leaders, March-April 2004. The Kurdish leadership has proven willing to compromise on a number of issues. For example, regarding the status of Kirkuk, the Kurds agreed to postpone a final resolution of this issue rather than insisting on its immediate incorporation into the Kurdish region. However, it is simply inconceivable that the Kurdish leadership could compromise on the issue of a separate Kurdish region within a federal Iraq.

(36) Bicommunal federations have a particularly poor track record in terms of survival. On the problems associated with bicommunal federations, see Ronald Watts, Comparing Federal Systems, pp. 113-114; Ivo Duchacek, "Dyadic Federations and Confederations," Publius: The Journal of Federalism 18 (1988): 5-31.

(37) Author interviews with Kurdish political leaders, March-April 2004.

(38) As Brendan O'Leary, who was directly involved in the drafting of the TAL, notes, the goal of the wording was to recognize officially the status of the Kurdish region while maintaining the pretense that ethnicity is not the defining feature of the region. Insofar as other ethnicities (Turkomans, Assyrians, Arabs) inhabit the Kurdish-controlled region, it is obviously not a monoethnic region, but the Kurds are heavily numerically dominant throughout and plainly view this as "their territory" (Brendan O'Leary, personal communication).

(39) For a critical analysis of the arguments against asymmetrical federalism, see John McGarry, "Asymmetrical Federalism and the Plurinational State" (working draft position paper for the Third International Conference on Federalism, Brussels, 3-5 March 2005).

(40) Of course, this formulation can serve more than one political purpose. After decades of perceived oppression at the hands of Sunni regimes, it may simply reflect a desire on the part of the Shi'a to move Iraq beyond ethnic/sectarian identities. At the same time, however, as the majority community in Iraq, it is the Shi'a who benefit politically by treating all groups equally (i.e., not granting special rights or protections to minority groups).

(41) The full text of the TAL available at http://www.cpa_iraq.org/ government/TAL.html.

(42) Under the system outlined in the TAL (or something similar), this basic problem of constitutional asymmetry would be exacerbated by the bicommunal nature of the resulting federation. On the problems of resolving political issues in bicommunal federations, see David E. Schmitt, "Problems of Accommodation in Bicommunal Societies," Conflict Quarterly 11 (Fall 1991): 7-18.

(43) For the text and a defense of al-Rubaie's proposal (which he terms "Democratic Regionalism"), see http://institutkurde.org/confIrak/Rubaie.htm (accessed 20 September 2005).

(44) On the evolution of the Spanish model, see Robert Agranoff, "Federal Evolution in Spain," International Political Science Review 17 (1996): 385-401.

(45) Christopher Dickey, "Make or Break," Newsweek, 10 November 2004; http:/msnbc.msn.com/id/6453400/site/newsweek/print/1/.

(46) Khaled Yacoub Oweis, "Interview--Iraq's Chalabi Suggests Autonomy for Oil-Rich South," Yahoo! India News, 28 January 2005; http://in.news.yahoo.com/050127/137/ 2j9hb.html.

(47) See Institute of War and Peace Reporting, "Federalism Demands in South," IWPR's Iraqi Crisis Report 111 (4 February 2005); "Sunni Powers Oppose Federalism in Western Iraq," Assyrian International News Agency, 24 March 2005; http://www.aina.org/news/20050324135407.htm.

(48) Those apparently opposed to the idea include prominent Sunni Arab groups the AMS and the IIP, along with Muqtada al-Sadr's faction. However, these groups oppose any form of federal system whether on the basis of eighteen provinces or five regions. The other major force in Iraqi politics that apparently (and perhaps fatally) opposes the five-region model is the United States (see Christopher Dickey, "Make or Break").

(49) For example, the Mutada al-Sadr's Mahdi Army has clashed violently with SCIRI's Badr Brigade on the streets of Karbala and Najaf on several occasions.

(50) The support for religious parties as evidenced by the recent elections is almost certainly exaggerated, however. In particular, the city/province of Basra has a long tradition of secularism and tolerance. The absence of viable secular alternatives at present and Sistani's tacit endorsement of the UIA (and, thereby, its constituent parties) undoubtedly helped the cause of religious parties.

(51) The Iraqi Governing Council, a body appointed by Paul Bremer, head of the Provisional Coalition Authority, was an almost perfect microcosm of Iraq's various divisions. The twenty-five members of the council comprised thirteen Shi'a Arabs (a mix of religious and secular), five Sunni Arabs, five Kurds, an Assyrian, and a Turkoman.

(52) The Interim Government (also a U.S. creation) was made up of Iyad Alawi (secular Shi'a, prime minister), Ghazi al-Yawer (Sunni Arab, president), Ibrahim al-Jaafari (religious Shi'a, vice president), and Rowsch Shaways (Kurd, vice president).

(53) One of the most notable features of the recent election was the almost total absence of crossover parties with an appeal capable of transcending communities. In general, communities, whether Kurds, Turkomans, Sunnis, Shi'a or Assyrian, ran and voted for their own community parties.

(54) See, for example, Jim Michaels and Rick Jervis, "Dispute Escalates between Shiites, Sunnis," USA Today, 18 April 2005; http://wwwusatoday.com/news/world/iraq/ 2005-04-18-shiites-sunnis_x.htm; Solomon Moore, "Recent Violence Stirs Sectarian Tensions"; Thanassis Cambanis, "Fractured Iraq Sees a Sunni Call to Arms," The Boston Globe, 27 March 2005; http://www. boston.com/news/world/middleeast/articles/2; Nicholas Blandford, "Ethnic Divide Deepens in New Iraq," The Christian Science Monitor, 8 March 2004; http://www.csmonitor.com/2004/0308/p01s04-woiq.htm.

(55) On increasing levels of ethnic tension in Mosul and Kirkuk, see Patrick Cockburn, "Iraq's Northern Capital Stalked by Suicide Bombers," The Independent, 15 April 2005; http://news.independent.co.uk/low_res/story.jsp?story=6; Soran Dawoodi, "Rallies Provoke Fears of Unrest in the Ethnically-Mixed City of Kirkuk," KurdishMedia, 11 October 2004; http://www.kurdmedia.com/ news.asp?id=5583; Mark McDonald, "If Civil War Erupts after Handover, Kirkuk May Be Its Starting Point," Knight Ridder Newspapers, 24 June 2004; http://www.realcities.com/mld/krwashington/ 9004241.htm?template=contentModules/prin; George Packer, "The Next Iraqi War," The New Yorker, 7 October 2004; http://www.newyorker.com/printable/?fact/041004fa_fact; Steve Fainaru, "In Mosul, a Battle 'Beyond Ruthless,'" The Washington Post, 13 April 2005; http://www.washingtonpost.com /ac2/wp-dyn/A48017-2.

(56) Of course, the design of Iraq's federal system is only one among many institutional choices that will have a bearing on the coherence of Iraq's political space. Other institutional choices--the use of a proportional representation electoral system, for example--will also be a major factor in this equation.

Table 1
Results of elections to the Iraqi National Assembly, 30 January 2005

 Total Vote
Electoral alliance votes percentage Seats

United Iraqi Alliance 4,075,292 48.19 140
Kurdish Alliance 2,175,551 25.73 75
Iraqi List 1,168,943 13.82 40
The Iraqis 150,680 1.78 5
Iraqi Turkmen Front 93,480 1.11 3
National Independent Cadres and
 Elites (Shi'a, religious) 69,938 0.83 3
People's Union (Communist) 69,920 0.83 2
Islamic Group of Kurdistan 60,592 0.72 2
Islamic Action Organization
 (Shi'a, religious) 43,205 0.51 2
National Democratic Alliance 36,795 0.44 1
National Rafidain List (Assyrian) 36,255 0.43 1
Reconciliation and Liberation Bloc 30,796 0.36 1
Others 444,819 5.65 0
Total 8,456,266 100.00 275

Source: "Iraqi Legislative Election Results, 2005"; http://
en.wikipedia.org/wiki/Iraqi_legislative_election_results%2C_2005.

Article Details

Printer friendly Cite/link Email Feedback
Author:	Anderson, Liam; Stansfield, Gareth
Publication:	Publius
Geographic Code:	7IRAQ
Date:	Jun 22, 2005
Words:	10619
Previous Article:	Public opinion on issues of U.S. federalism in 2005: end of the post-2001 pro-federal surge?
Next Article:	The Nigerian federation at the crossroads: the way forward.
Topics:	Elections Forecasts and trends Political aspects Federalism Analysis Meritocracy Political systems Evaluation